Whether IQ tests measure intelligence depends on how you define "intelligence". IQ tests measure ability to solve visual and/or verbal puzzles, and it's reasona

Whether IQ tests measure intelligence depends on how you define "intelligence". IQ tests measure ability to solve visual and/or verbal puzzles, and it's reasonable to call that a type of intelligence. I'm not saying that's the only form of intelligence, but it's clear that IQ scores have _something_ to do with intelligence. But that question is ultimately a red herring. IQ scores are partially biologically heritable. IQ scores are correlated with job performance. So there exists a biologically heritable factor that's correlated with job performance. That's very relevant for discussions about social mobility, whether or not you want to call that factor "intelligence". You asked for a citation. The strongest evidence for the claim that IQ is biologically heritable comes from studies of twins. A meta-analysis of 111 separate studies (https://sci-hub.la/10.1126/science.7195071 https://sci-hub.la/10.1126/science.7195071) found that when identical twins were separated and raised in different families, their IQs were correlated at 0.72. Fraternal twins raised together had IQs correlated at only 0.60. Biologically unrelated children raised together (an adopted child and a biological child) had IQs correlated at 0.29. (Side note: Before you demand someone provide a citation, try googling it for yourself. This result isn't hard to find.) You have a good point that the mechanism could be epigenetic rather than genetic. Here's a study that found a hint of a link between epigentic factors and intelligence: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2892514/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2892514/. Here's a study finding that epigenetics are correlated strongly between identical twins: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3063335/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3063335/ On the other hand, the epigenetic correlation decreases as the twin pairs get older, but the IQ correlation actually increases as twin pairs get older. So it's not clear. It's also worth noting that IQ correlates with income at only 0.23 according to http://www.emilkirkegaard.dk/en/wp-content/uploads/Intelligence-and-socioeconomic-success-A-meta-analytic-review-of-longitudinal-research.pdf http://www.emilkirkegaard.dk/en/wp-content/uploads/Intellige.... In general, I agree it's difficult to draw unambiguous conclusions from the data we have. I think the data are consistent with the possibility that genetic inheritance of intelligence plays a major role in reducing social mobility, and also consistent with the opposite possibility. But you are demanding a much higher standard of evidence for the claim that intelligence is genetic than for the claim that it isn't, and I don't think that's fair.